Advances in HAART have provided us with some surprising insights into the molecular events that underlie the pathogenesis and neurologic disease that accompanies HIV-1 infection of the CNS. In particular, significant amelioration of pre-existing neurologic disease in patients with HAART and lack of correlation between numbers of apoptotic neurons and pre-mortem neurologic status suggests that HIV-1 associated neurologic disease may have a component of reversible metabolic encephalopathy. Accurately modeling this in vitro requires different assays than measuring neuronal apoptosis as an endpoint after exposure to HIV-1 neurotoxins. We hypothesize that initial toxicity may occur in nerve terminals exposed to HIV-1 neurotoxins causing the terminal to undergo oxidative stress, mitochondrial dysfunction, with resultant impairment of the terminal to undergo synaptogenesis or maintain normal synaptic communication by synaptic vesicle recycling. To successfully model these events, we need to employ a compartmentalized culture that separates neuronal cell bodies from their distal neurites and synapses by means of a diffusion barrier and geographically defined tracks that allow neurite migration and synapse formation and function to be quantified. In this R21 application, we will investigate the responses of cortical neurons, explanted with astrocytes into compartmentalized cultures to undergo process (i.e. neurite) outgrowth, express markers for synapses, and establish metabolic activity in pre-synaptic nerve terminals after exposure to the HIV-1 neurotoxins PAF and Tat. These neurotoxins have the relatively unique ability to up-regulate glycogen synthase kinase 3 beta (GSK-3beta), an enzyme important in modulating neuronal cytoskeletal events that may be critical for neuritogenesis and synaptogenesis. The Specific Aims are: 1) To establish conditions for quantitative assays of neurite outgrowth, reactive oxygen production and mitochondrial membrane potential in situ, and for synaptic vesicle recycling in neurites separated by a diffusion barrier from their cell bodies; 2) To investigate the effects of Tat and PAF on these parameters; and 3) To investigate the relative contribution of GSK-3beta by using viral vectors that can deliver in situ constitutively active or inactive forms of GSK mutants to either neuronal cell bodies or neurites in different compartments. Results from these experiments will significantly advance our understanding of how vulnerable neurons respond to HIV-1 neurotoxicants prior to apoptosis.